基于海量遥感仿真数据的图像缓存置换算法改进及应用  

Improvement and Application of Image Cache Replacement Algorithm Based on Massive Remote Sensing Simulation Data

作　　者：袁帅 YUAN Shuai(School of Henan University Minsheng College,Kaifeng Henan 475004)

机构地区：[1]河南大学民生学院,河南开封475004

出　　处：《数字技术与应用》2021年第9期13-15,共3页Digital Technology & Application

基　　金：面向智能网联汽车感知模块的多模态融合目标检测研究(212102210544)。

摘　　要：在进行高分系列卫星的模拟仿真时,由于其超大视场成像、高时空分标率、单景数据量巨大的特点,用户需进行长时间、大范围、多条CCD拼接等情况的遥感仿真。高频度的进行此种仿真,这在以往的光学仿真中是很难见到的,所以运用传统仿真运算模型时,会产生内存溢出和导致电脑"卡顿",严重影响了仿真运算效率。为了更高效的进行遥感仿真,本文设计了一种新型的瓦片动态加载策略,用来有效缓解遥感仿真中瓦片加载内存溢出现象。In the simulation of high resolution satellite series,due to its super large field of view imaging,high space-time scaling rate and huge amount of single scene data,users need to carry out remote sensing simulation of long-time,large-scale and multi CCD splicing.This kind of simulation at high frequency is difficult to see in the previous optical simulation,so when using the traditional simulation model,it will cause memory overflow and computer"stuck",which seriously affects the efficiency of simulation.In order to carry out remote sensing simulation more efficiently,this paper designs a new tile dynamic loading strategy,which can effectively alleviate the tile loading memory overflow phenomenon in remote sensing simulation.

关 键 词：遥感仿真 海量信息存取 并行计算 算法改进 

分 类 号：TP75[自动化与计算机技术—检测技术与自动化装置]